Location of the segmental mandibular defects on an explanted mandible on the buccal (A) and lingual (B) surfaces.
Plating patterns used for the 0- and 1-mm (A), 3-mm (B), and 5-mm (C) defects.
Intraoperative view of a 5-mm defect stabilized with internal fixation.
Scores are described in the Micro-CT (micro–computed tomography) Analysis subsection of the Methods section. Error bars indicate 95% confidence intervals. For pairwise relationships on Mann-Whitney testing, P < .05 (0 vs 3 mm, 0 vs 5 mm, 1 vs 3 mm, 1 vs 5 mm, and 3 vs 5 mm).aAll rats with 5-mm defects received a mean bone union score of 0.
Scores are described in the Micro-CT (micro–computed tomography) Analysis subsection of the Methods section. Error bars indicate 95% confidence intervals. For pairwise relationships on Mann-Whitney testing, P < .05 (0 vs 3 mm, 0 vs 5 mm, 1 vs 3 mm, 1 vs 5 mm, and 3 vs 5 mm).
Three-dimensional reconstructions of representative segmental defects from each group are depicted.
DeConde AS, Lee MK, Sidell D, Aghaloo T, Lee M, Tetradis S, Low K, Elashoff D, Grogan T, Sepahdari AR, St John M. Defining the Critical-Sized Defect in a Rat Segmental Mandibulectomy Model. JAMA Otolaryngol Head Neck Surg. 2014;140(1):58-65. doi:10.1001/jamaoto.2013.5669
Advances in tissue engineering offer potential alternatives to current mandibular reconstructive techniques; however, before clinical translation of this technology, a relevant animal model must be used to validate possible interventions.
To establish the critical-sized segmental mandibular defect that does not heal spontaneously in the rat mandible.
Design and Setting
Prospective study of mandibular defect healing in 29 Sprague-Dawley rats in an animal laboratory.
The rats underwent creation of 1 of 4 segmental mandibular defects measuring 0, 1, 3, and 5 mm. All mandibular wounds were internally fixated with 1-mm microplates and screws and allowed to heal for 12 weeks, after which the animals were killed humanely.
Main Outcomes and Measures
Analysis with micro–computed tomography of bony union and formation graded on semiquantitative scales.
Seven animals were included in each experimental group. No 5-mm segmental defects successfully developed bony union, whereas all 0- and 1-mm defects had continuous bony growth across the original defect on micro–computed tomography. Three of the 3-mm defects had bony continuity, and 3 had no healing of the bony wound. Bone union scores were significantly lower for the 5-mm defects compared with the 0-, 1-, and 3-mm defects (P < .01).
Conclusions and Relevance
The rat segmental mandible model cannot heal a 5-mm segmental mandibular defect. Successful healing of 0-, 1-, and 3-mm defects confirms adequate stabilization of bony wounds with internal fixation with 1-mm microplates. The rat segmental mandibular critical-sized defect provides a clinically relevant testing ground for translatable mandibular tissue engineering efforts.
Defects in the mandible can arise from a variety of benign and malignant processes.1,2 At present, the preferred method of reconstruction of segmental mandibular defects uses 1 of several potential vascularized free flaps.3 Although free tissue transfer is a reliable means of mandibular reconstruction, osseous flaps frequently require prolonged operations and carry a 20.5% risk of perioperative medical complications.4,5 To date, several investigations into alternative methods of mandibular reconstruction have been reported. In addition to animal models, the off-label use of growth factor–impregnated scaffolds for the reconstruction of segmental defects that span the entire height of the mandible (through-and-through segmental mandibular defects) has also been described in humans.6,7 However, before wide-scale translation of innovations in tissue engineering to humans, new technologies will require robust data from animal models.
A variety of large animal species have been used to investigate segmental mandibular regeneration. The canine model is the most common segmental mandibular defect used for this purpose,8- 11 although goat, sheep, cat, rabbit, and monkey models have also been described.12- 16 Although these large-animal models provide a more relevant representation of the volume- and load-bearing conditions of the defects seen in humans, they also have comparatively more sentience and more ability to suffer, are more expensive, and are protected under the Laboratory Animal Welfare Act. A small-animal segmental mandibular defect model would allow for improved control of strain, sex, age, weight, and husbandry conditions while adhering to the principles of laboratory animal use in regard to reduction, replacement, and refinement.17
When considering mandibular defect models, small animals have been effectively used to investigate distraction osteogenesis.18 However, several unique characteristics of the distraction osteogenesis model merit mention because these characteristics limit the utility of this model as a platform to investigate tumor resection defects. In the distraction model, animals develop severe malocclusion, fundamentally altering the forces of mastication. Similarly, external fixation is required despite its rare application after segmental mandibular resection in clinical conditions. In a resection model, a mismatch occurs in the proximal and distal mandible bony edges, whereas in a distraction model, a single osteotomy is performed and thus a perfect match results. An end-to-end mismatch is an important consideration from a functional standpoint but will also physiologically alter the spatial distribution and volume of growth factors and cytokines critical for mandibular regeneration.
In concert, these fundamental differences have become the impetus for the present investigation. The purpose of this report is to continue our pursuit of a small-animal model that better represents the anatomic and physiologic qualities of a human segmental mandibular wound, thus providing a new platform for future evaluation of novel mandibular regeneration interventions. A previous study has described a reliable small-animal model that can be used for the study of composite mandibular defects.19 The present study seeks to advance the utility of this animal model by determining the critical-sized through-and-through defect of the mandible that will not heal spontaneously.
Twenty-nine male 4-month-old Sprague-Dawley rats were divided into 4 groups of increasing defect size to establish the segmental critical-sized mandibular defect in a small animal. Critical-sized defects of the mandible are defined as those that will not heal spontaneously during the expected natural life of the animal. Surgical treatments consisted of creation of 1-, 3-, and 5-mm segmental mandible defects with internal rigid fixation with 1.0-mm microplates (Synthes and Stryker Corporation). In addition, a 0-mm defect was created to confirm that internal fixation was adequate to allow primary bone healing in the event that the 1-mm defect did not form a bony union. Because a 1-mm cutting burr was used for the 0-mm defects, these mandibles required plating after reducing the 1-mm kerf so that primary bone healing could occur (explained below). Animals were allowed to heal for 12 weeks, and then mandibular regeneration was evaluated by micro–computed tomography (micro-CT).
All animal protocols were approved and overseen by the animal research committee at the University of California, Los Angeles (UCLA). The UCLA facility is accredited by the Association for Assessment and Accreditation of Laboratory Animals. This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. Innate bone-healing abilities were tested in 12-week-old Sprague-Dawley rats as previously described.19 In brief, the animals received inhalational isoflurane until deep anesthesia was achieved. The animals were then shaved on the ventral surface of the mandible and prepared and draped in a sterile fashion. Using aseptic technique, an incision overlying and parallel to the left mandible was carried down through subcutaneous tissues. The inferior border of the mandible was then exposed by dividing the pterygomasseteric sling with electrocautery. The lingual and buccal surfaces of the mandible were then exposed through a supraperiosteal elevation of the musculature. Using a 1-mm high-speed cutting burr at 3000 rpm with copious irrigation, defects were created (Figure 1). Hemostasis was then achieved with electrocautery, and the wound was irrigated free of bone dust. The appropriate plate was applied (Figure 2) to create the desired segmental defect. The pterygomasseteric sling was then reapproximated with resorbable suture, and the skin was closed with nonresorbable nylon sutures. Rats were then allowed to recover from anesthesia and transferred to the vivarium for postoperative monitoring. Postoperatively, all animals received analgesia with subcutaneous injections of buprenorphine hydrochloride (0.1 mg/kg) for 72 hours postoperatively. All animals received a combination of trimethoprim and sulfamethoxazole in the water supply for 1 week after the operation as prophylaxis against infection. Animals were fed a gel diet (S5769 Nutra-gel Purified Formula; Bio-Serv) for the first week, followed by a soft diet (F3580-1 Bacon Softies; Bio-Serv) for the rest of the study period. Weights were measured on a weekly basis to assess nutritional status. Animals were humanely killed 12 weeks postoperatively to allow for adequate healing time. Preliminary studies of a marginal-defect Sprague-Dawley rat model demonstrated that healing required 8 weeks.20 Prior studies in the distraction osteogenesis literature demonstrate healing of the distraction defect at 4 weeks.18 By allowing for a conservative 12-week healing period, the larger segmental defect was ensured adequate opportunity to form a bony union.
The 0-mm defect was created by reducing the kerf created by the single pass of a 1-mm cutting burr. A 4-hole 1-mm box plate was predrilled only on the caudal side of the anticipated burr cut. The mandible was stabilized with an Adson forceps by grasping the bicortical cranial portion of the mandible while a single pass of the cutting burr that traveled from the inferior border to the sigmoid notch was made. The nadir (ie, the most ventral extension) of the sigmoid notch was used as a landmark for creation of the most caudal cuts (Figure 1). This intraoperative landmark is relatively consistent and easily recognizable. In addition, the monocortical ramus is structurally weak; in prestudy cadaveric trials, we found that the thin monocortical bone frequently fractured along craze lines created by the high-speed cutting burr if placed posterior to the nadir of the sigmoid notch. After the single pass was made with the cutting burr, the box plate was then applied to the buccal aspect of the ramus of the mandible with 1-mm screws. The 1-mm kerf loss from the width of the drill bit was reduced, ensuring direct apposition of the bony edges before plating, thus creating the 0-mm defect. While maintaining the mandible in reduction, the rostral screw holes were drilled through the box plate and secured with 1-mm screws (Figure 2A).
The 1-mm defect was created by predrilling the buccal side of the mandible for a 4-hole 1-mm box plate. The kerf loss from a single pass of the 1-mm cutting burr created the defect. The plate was then applied with four 1-mm screws to the buccal aspect of the mandible (Figure 2A).
The 4-hole box plate cannot span the 3- and 5-mm defects. A prebent 6-hole box plate was used for the 3-mm defects (Figure 2B). The 5-mm defect cannot be plated in the same way because the body of the mandible does not have enough height to receive both screws of a box plate. The 5-mm defect was double-plated along the dorsal and ventral borders of the mandible. A prebent ventral plate with 2 holes on either side of the fracture line stabilizes the fracture to intraoperative palpation. A second ventral plate acts as a tension band, as in human mandibular fixation, further stabilizing the mandible in an additional plane as the box plates do (Figure 2C and Figure 3).
The 3- and 5-mm defect mandibles were predrilled before making the cuts. The anticipated defects were measured and marked. Defects were created with 2 passes of the 1-mm cutting burr. The caudal cut was always made first and always began at the nadir of the sigmoid notch. The caudal cut was made first to allow grasping and stabilization of the mandible on the stronger bicortical dentate region. The predrilled plates were then applied, and a machined sizing block was used to confirm accurate creation of the 3- and 5-mm defects.
The animals were killed after 12 weeks, and the mandibles were harvested for analysis. Specimens were fixed in 10% formalin for 48 hours and rinsed in phosphate-buffered saline solution before imaging with high-resolution micro-CT (micro-CT 40; Scanco USA, Inc). Micro-CT data were collected at 50 kV (peak) and 160 μA. Visualization and reconstruction of the data were performed using orthodontic software (Dolphin 3D; Dolphin Imaging & Management Solutions). Three-dimensional surface reconstructions of the mandible, including the regions of interest, were made using the entire volumetric data set. The scatter artifact from the titanium microplates precluded quantitative analysis of the scans because artifact cannot be reliably distinguished from newly formed bone by the current software. Therefore, 3 of us (A.S.D., M.K.L., and A.R.S.) reviewed the images independently and scored the scans as previously described.21 The examiners, blinded to the identity of the specimens, were allowed to use the software’s dynamic segmentation function to create the most realistic appearance of the mandible with minimal loss of unwounded cortical bone owing to thin structures and minimal superimposition of artifacts and soft tissue. In short, the defects were graded on 2 scales. Bony union was graded on a scale from 0 to 3, representing no bony bridging (score of 0), bridging of less than half (1), bridging of greater than half (2), and complete bridging of the bony gap by newly formed bone (3). Bony formation, a determination of the volume of new bone growth within the bony defect, was graded on a scale from 0 to 4, where 0 represents no new bony volume and 4 represents regrowth of greater than 95% of the original defect.
To assess the pairwise agreement of raters, we computed the percentage of agreement and the κ statistic. To assess the agreement among the set of 3 observers, we computed the multirater agreement κ statistic (using the MKAPPASC macro in SPSS [SPSS, Inc]).22 Each rat was measured 3 times for bony formation and union. The median score of those 3 measurements was used for the analysis. To evaluate overall differences among the 4 defect groups (0-, 1-, 3-, and 5-mm), we performed a Kruskal-Wallis test. If the Kruskal-Wallis test result was statistically significant, pairwise differences among the 4 groups were tested with the Mann-Whitney test. Comparisons among groups were deemed statistically significant at the threshold of α < .05 threshold. Statistical analyses and plots were performed using commercially available software (R, version 2.15.0 [http://cran.r-project.org/bin/windows/base/] and SPSS, version 19 [SPSS, Inc]).
A total of 29 rats underwent surgery for a segmental mandibulectomy. Three rats (10%) experienced complications that precluded inclusion in the analysis. One intraoperative death occurred secondary to exsanguination from an arterial injury from the cutting burr during creation of a 3-mm defect, and therefore this animal was replaced with another. Any screws that were stripped or did not adequately secure the plates were replaced with rescue screws until the mandible was palpably stable across the defect. In addition, overtightening 1 rescue screw in the ramus of the mandible of a rat with a 5-mm defect split the mandible along a craze line. This rat was killed humanely because the fractured mandible could not be repaired. No other hardware failures (ie, fractured plates, extruded plates or screws) occurred. One rat with a 1-mm defect developed stridor on postoperative day 0. Aspiration of wood chips has been previously described19; therefore, direct laryngoscopy was performed with a freer elevator after reinduction of anesthesia with isoflurane. Wood chips were indeed identified in the oropharynx and removed with forceps. The rat awakened with quiet respiration, and otherwise had an uncomplicated recovery. One rat with a 3-mm defect developed an abscess that presented on postoperative day 7. This animal was killed humanely on discovery of the abscess. The rats were then allowed to heal for 12 weeks, at which point they were killed humanely. All animals had some degree of weight loss during the first postoperative week (mean weight loss, 3% and always <10%), after which they gained weight appropriately by age. In total, 7 rats underwent analysis in the 0- and 1-mm defect groups, and 6 rats underwent analysis in the 3- and 5-mm defect groups.
Interobserver variability for the bone union and formation scores was excellent among the 3 clinicians. Pairwise comparison of reviewers demonstrated excellent agreement on bony union (reviewers 1 vs 2, κ = 0.95; reviewers 1 vs 3, κ = 0.84; reviewers 2 vs 3, κ = 0.79) and very good agreement on bony formation (reviewers 1 vs 2, κ = 0.89; reviewers 1 vs 3, κ = 0.68; reviewers 2 vs 3, κ = 0.58). Similarly, multirater agreement was excellent for bony union (κ = 0.86) and very good for bony formation (κ = 0.72).
Figure 4 and Figure 5 summarize the radiographic semiquantitative scoring. Figure 6 demonstrates 3-dimensional reconstruction of micro-CT scans from representative samples in each experimental group. The Kruskal-Wallis 1-way analysis by ranks revealed that bony formation and bony union varied with defect size. Significant (P < .05) pairwise relationships by Mann-Whitney testing are similarly displayed in Figures 4 and 5. Most important, none of the 6 rats that underwent 5-mm segmental mandibulectomies had any bony bridging, and the 3-mm defects all had some degree of bony bridging.
Through-and-through mandibular defects can arise from a variety of pathological processes, including congenital defects, neoplasm, infection, osteoradionecrosis, and trauma. Failure to restore mandibular continuity results in devastating consequences for patients, including oral incompetence, impaired mastication, and facial contour defects.5,23,24 The current standard of care for reconstruction of large segmental defects is autologous tissue transfer in the form of microvascular anastomosis of bony free flaps.3 Although microvascular flap reconstruction is a surgically reliable means of reconstructing these defects, it requires specialized training, is expensive,3 and brings significant medical morbidity with prolonged operative times.4,25 These limitations are the spur for investigations of novel tissue-engineering approaches, and a variety of novel growth factors, cellular therapies, and biomimetic scaffolds hold the promise of obviating autologous tissue transfer in the future.15,26,27 However, before translation of new technologies to clinical practice, experiments using large-scale animals are required to establish safety, reliability, and effectiveness of these novel therapies.28,29 The small-animal model optimizes valuable research resources by providing clinically relevant data with reasonable work, time, and cost.
A reproducible segmental bony wound that does not heal spontaneously in a small animal would offer an ideal platform to test novel mandibular regeneration interventions, such as scaffolds and cellular therapies. All of the 5-mm defects were unable to form a bony union spontaneously on micro-CT, whereas 0-, 1-, and 3-mm defects had some degree of bony union and formation present (Figures 4-6). Future studies investigating regenerative techniques should use the 5-mm defect to see if any healing can be induced experimentally. Any smaller defect may experience innate spontaneous bony healing of the wound regardless of the intervention being investigated, as occurred in the 3-mm defect in this study.
The small size of the laboratory rat makes surgical interventions technically difficult; however, the rat mandible has already successfully served as an animal model for partial-thickness critical-sized defects.30,31 The 2 mandibular critical-sized defects described in previous studies of the rat model are a 5-mm trephine posterior to the root of the incisor and a 5 × 5-mm square defect based on the inferior border also posterior to the root of the incisor.30- 32 The inherent stability of partial-thickness defects precludes placement of hardware for fixation and does not model the hardware-related challenges encountered clinically in segmental defects. Sidell et al19 described the first segmental resection defect in a rat mandible but did not establish the critical-sized defect that is essential to investigations of mandibular healing and regeneration. In addition, a proven means of reliable fixation allowing for bony healing was not established, which is a significant technical hurdle in the small-animal model. The following 2 methods of internal fixation are described: a titanium internal fixation and a polypropylene splint fixated with Kirschner wires. Studies in segmental femoral defects in rats demonstrate that internal plate fixation is effective at resisting torsion and compression and is easier to apply than Kirschner wire–based devices.33 The ease and reproducibility of the internal titanium plate fixation is also true for the rat mandible and served as the reasoning behind using this device. Indeed, the present study establishes that internal plate fixation provides enough stability to heal non–critical-sized bony defects (0-, 1-, and 3-mm defects).
The distraction osteogenesis literature has already effectively implemented the mandibular distraction model in the rat.16 This animal model has allowed successful demonstration of unique vascularity34 and molecular expression differences35 in the healing of bony gaps and distraction. This work highlights the value of a small-animal model as well as the need for a unique segmental defect resection model. Distraction osteogenesis exploits the innate fracture-healing processes through incremental increase of an osteotomy. Immediate intraoperative distraction with external fixation has also demonstrated that rats are unable to heal a 5-mm defect without the development of a fibrous union. However, distraction osteogenesis is fundamentally different from mandibular resection. In the present study, mandible defects were the result of resection rather than distraction. As a result, occlusion was preserved postoperatively with internal fixation, thus maintaining an even distribution of the stresses of mastication on the remaining mandible. This result provides a stark contrast to the severe malocclusion that occurs in the setting of distraction osteogenesis as demonstrated by Buchman et al.18 In addition, a resection model recreates the mismatched bony edges (eg, a tall ramus and a shorter body) inherent to mandibular defects after tumor resection. Future investigations of mandibular regeneration will need to overcome the spatial mismatch that is frequently encountered after mandibular resection.
The semiquantitative analysis of the micro-CT imaging imposes some limitations on the analysis. The radiodensity evaluated by the observers is interpreted as bony regrowth, which is a reasonable assumption given that no scaffolds were used. In the case of a hydroxyapatite scaffold, the radiographic findings will need validation with a histologic analysis to determine that the radiodensity is indeed bony regrowth and not scaffold persistence.21 This validation can be a technically challenging feat in a specimen that has not formed a union given that even minimal dissection of soft tissue and sectioning of the sample can cause distortion of the region of interest, potentially biasing standardized histologic analysis.
The small scale of the laboratory rat affords many advantages as an animal model but also contributes to a technically challenging procedure. In the present study, 2 of 29 rats (7%) had technical and preventable surgical errors that led to the rats’ early deaths. Only 1 rat (3%) developed an abscess. Prior study of this animal model by Sidell et al19 had higher rates of abscess and/or fistula (8 of 72 [11%]), which may in part result from a lack of intraoral wounds in that study. Intraoral wounds theoretically increase the risk of infectious complications that would preclude analysis. Therefore, the decision was made to create bony wounds outside the oral cavity. Future studies should investigate an animal model that contains an intraoral wound (or should include intraoral wounds in the present animal model) in addition to defects in other parts of the mandible that have unique anatomic challenges (eg, the symphyseal defect).
We experienced 2 surgical complications: 1 rat exsanguinated intraoperatively and 1 rat mandible fractured along a craze line from overtightening of the screw. These complications illustrate the 2 most challenging aspects of this model: fitting adequate hardware on a small mandible and control of hemorrhage. Arterial bleeding from a branch of the facial artery, the supplementary mandibular artery19 that lies at the posterior aspect of the segmental defect, is not commonly encountered. Some anatomic variation exists in the mandibles, and the vessel is encountered in less than half of the surgical procedures. Arterial bleeding when encountered can typically be stopped with some combination of pressure, electrocautery, and suture ligation. However, the rats are small and cannot tolerate much beyond 1.5 mL of blood loss. Similarly, placement and securing of the plates require precision to accommodate enough screws in the mandible to allow for adequate fixation without fracturing the mandible. These challenges provide a learning curve for the surgeon; in our experience with this study, the surgeon averaged approximately 45 minutes per rat after the first 5 rats. Coupled with ease of anesthesia, recovery, and husbandry of a small animal, the rat model of the segmental mandibular defect is time efficient.
No perfect animal model exists. All small-animal models offer improved control over environment, husbandry, and biological variables at the cost of a compromise of the modeling of clinical forces and volumes,17 but the model still can serve as an ideal refinement ground for early bony regeneration interventions. In addition, small-animal models (eg, mice and rats) consist of a more primitive skeletal system that does not possess haversian canals and does not exactly mirror bony healing in humans.36 Therefore, the small-animal model is not designed to replace the large-animal model but to complement it. Findings in the small-animal model should also be confirmed in higher-level animals before human clinical trials.
The present study’s data establish the first critical-sized segmental mandibular defect in a small-animal model at 5 mm. In addition, the study confirms that internal fixation with titanium microplates across reproducible bony gaps is technically feasible and provides adequate fixation to allow development of bony union in defects smaller than 5 mm.
Segmental mandibular defects commonly arise from a variety of head and neck diseases and abnormalities. Current treatments using microvascular autologous tissue transfer are reliable but carry a significant risk of morbidity, and a search for alternative means of reconstruction using novel technologies is ongoing. An essential tool for understanding and manipulating bone regeneration is a critical-sized defect. To date, only large-animal segmental mandibular models have been investigated. The present study establishes the first critical-sized segmental mandibular resection defect in a small-animal model.
Submitted for Publication: April 7, 2013; final revision received August 17, 2013; accepted September 15, 2013.
Corresponding Authors: Tara Aghaloo, DDS, MD, PhD, Division of Diagnostic and Surgical Sciences, School of Dentistry, University of California, Los Angeles, Box 951668, 53-009 Den Bldg, Los Angeles, CA 90095-1668 (email@example.com); and Maie St John, MD, PhD, Department of Head and Neck Surgery, David Geffen School of Medicine, University of California, Los Angeles, 62-132 CHS, 10833 Le Conte Ave, Los Angeles, CA 90095-1624 (firstname.lastname@example.org).
Published Online: November 14, 2013. doi:10.1001/jamaoto.2013.5669.
Author Contributions: Dr DeConde had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: DeConde, Sidell, Aghaloo, M. Lee, Tetradis, St John.
Acquisition of data: DeConde, M. K. Lee, Aghaloo.
Analysis and interpretation of data: DeConde, M. K. Lee, M. Lee, Tetradis, Low, Elashoff, Grogan, Sepahdari, St John.
Drafting of the manuscript: DeConde, Low, Grogan, St John.
Critical revision of the manuscript for important intellectual content: DeConde, M. K. Lee, Sidell, Aghaloo, M. Lee, Tetradis, Elashoff, Sepahdari, St John.
Statistical analysis: Low, Elashoff, Grogan.
Obtained funding: DeConde, Sidell, Aghaloo, M. Lee, St John.
Administrative, technical, and material support: DeConde, Aghaloo, M. Lee, St John.
Study supervision: DeConde, M. K. Lee, Sidell, Aghaloo, Tetradis, Sepahdari, St John.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported by the International Association for Dental Research and the Academy of Osseointegration, a UCLA Jonsson Cancer Center Transdisciplinary Cancer Research Grant, and by grant P30AG028748 from the National Institute on Aging. Statistical analysis was funded by grants UL1RR033176 and UL1TR000124 from the UCLA Clinical and Translational Science Institute.
Role of the Sponsors: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute on Aging or the National Institutes of Health.
Previous Presentation: This study was presented at the American Academy of Otolaryngology–Head and Neck Surgery Foundation Annual Meeting; September 9, 2012; Washington, DC.